312 ALASKAN GLACIER STUDIES and Dartmouth Glaciers, the streams from which are still carrying debris into the fiord. The eastern part of the fiord near these glaciers, however, is very shallow and the abundant reefs rising to the surface here are in some cases rock ledges, proving that there has not been such profound erosion as along the western side and also that there have not been extensive glacial deposits above the surface. The greater glacial scouring was apparently on the western side of the portion of College Fiord where the main current of the expanded trunk glacier was probably pushed by the eastern tributaries. Deposits might also have been made here by the Amherst and other tributaries before they retreated across to the shallow eastern portion of the fiord. A perfectly rational hypothesis to account for all the irregularities in the bottom of College Fiord would be to attribute them to glacial erosion without significant deposition. All of the existing basins and swells might be explained by a rhythmic alternation of plunging and rising related to the incoming of tributary ice streams. There is a basin below Harvard Glacier terminus and a swell not far above College Point near Smith Glacier. There is a basin south of College Point where the expanded Yale Glacier entered, below which is the swell west of Coghill River. Below this the water deepens nearly to Point Pakenham, perhaps under the influence of the incoming Dartmouth, Amherst, and adjacent glaciers. The places of relatively shoal water at the southern end of College Fiord, discussed in the next chapter, seem to be clearly of the moraine bar type, though we have no clear proof that they are not built upon a broad rock swell, just above the entrance of the former incoming Barry Arm Glacier. Besides the visible hanging valleys above tide water (PI. CXXV) there are several submerged hanging valleys (PI. CXXVI). The cove in which Wellesley Glacier now ends, and which has a maximum depth of about 48 feet, hangs 756 feet above the bottom of the main fiord. The arm of the fiord terminated by Yale Glacier hangs 498 feet above the main fiord (Fig. 42). This submerged hanging valley shows clearly that Harvard Glacier has always been the chief tributary of the College Fiord ice tongue, eroding more efficiently than the Yale Glacier at the time when the latter was a tributary of the former. The cove at the mouth of Coghill River probably has a similar hanging relationship; and College Fiord hangs above Port Wells, but how much is unknown because of the complication of glacial deposits. Glacial Deposits. Deposits of glacial origin are not common in the College Fiord region. They include (1) ground moraine, (2) terminal moraines, (8) lateral moraines and terraces, (4) outwash gravels, and (5) deposits below sea level. The ground moraine is everywhere thin and except for scattered erratic bowlders is entirely absent on steep slopes, but on the more gently sloping portions of the fiord walls, the deposit is thicker. The low flat strip of land on the southern side of Yale Arm has a thin veneer of glacial deposits, there being well developed morainic topography with ridges and hollows, the latter containing pools. Since the larger glaciers end in the water, terminal moraines are few in number. Mention has already been made of the long, narrow, piedmont strip on the western side of College Fiord, built up, in part at least, by deposits from the several cascading glaciers, such as the crescentic terminal moraines south of Bryn Mawr and Wellesley Glaciers. Possibly this area also includes some deposits of the earlier, expanded Harvard Glacier. Mendenhall states x that in the valley of Coghill River there are "several small lakes iMendenhall, W. C.. 20th Ann. Kept., U. S. Geol. Survey, Part VII, 1900, p. £73.